Thoracic irradiation is a major component of therapy for many patients with tumors of the breast, lung, esophagus, thymus and mediastinal lymphatics. In these patients, radiation-induced pulmonary injury is one of the most common treatment-related toxicities. Despite the large number of patients that receive thoracic irradiation, the biologic and physical determinants of radiation-induced lung injury are not understood fully. There are currently no means of predicting, prior to delivery of radiotherapy, what the physiologic consequences of the radiation will be. Previous studies in this area have been hampered by uncertainties regarding the three-dimensional (3-D) radiation dose distribution, lack of understanding of the mechanisms underlying the pathophysiology, use of inconsistent assessments of pulmonary injury, and confusion relating regional versus whole lung injury. While regional injury within the irradiated lung is almost universally seen, whole organ dysfunction is less common. This study will address the physical factors related to the development of radiation-induced lung dysfunction. The long-term goal of this project is to develop algorithms to predict, prior to radiotherapy, the patient's risk of developing symptomatic lung injury and the reduction in pulmonary function that the patient will experience. Advances in functional lung imaging and 3-D radiation treatment planning will be exploited to better understand the physical determinants of radiation- induced lung injury. Regional lung function will be assessed by single photon emission computed tomography (SPECT) lung perfusion scans. Changes in regional lung function will be correlated with regional radiation dose. Whole lung function will be assessed with pulmonary function tests and clinical symptoms. We believe that it is possible to directly relate changes in whole lung function to the following physical parameters: I.) the full 3-D dose distribution; 2.) the summation of regional lung injuries; 3.) baseline whole lung function; and 4.) preradiation functional heterogeneity within the lung. We hypothesize that changes in whole lung function result from the summation of dose-dependent regional injuries and the compensatory capacity of the functioning lung. We postulate that patients with good pulmonary reserve have greater compensatory abilities and are therefore less likely to develop whole lung dysfunction. This relationship between baseline pulmonary function, regional and whole lung injury will be explicitly addressed. An improved understanding of the physical determinants of radiation-induced lung injury and the development of predictive algorithms will improve patient care by reducing the incidence of clinically significant lung injury and facilitate "safe" dose escalation for intrathoracic tumors. While not part of this study, patients enrolled onto this protocol will be included in parallel biological studies designed to consider the molecular biological mechanisms of radiation-induced lung injury.